Pneumatic Brake Self-Diagnosis Using Learned Reaction References
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current diagnostic methods for electronically controllable pneumatic braking systems in commercial vehicles, especially in partially or fully autonomous vehicles, lack effective means for proactive fault detection and correction, particularly for manufacturing and age-related variances, and cannot account for subtle system reactions that do not affect functionality but may impact performance.
Innovation Solution
A diagnostic method that involves putting the braking system into a learning mode, performing predetermined activities while stationary or moving, detecting system reactions using existing sensors, and storing these reactions as references for subsequent tests, allowing for comparison and fault detection based on predefined limit values, and outputting fault signals to prevent automated mode activation if deviations are detected.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If driver's diagnostic capability is relied upon for fault detection, then the system can operate with simple diagnostic means, but the system cannot detect faults proactively in autonomous vehicles without a driver
Solution Approach 1:
The system performs preliminary diagnostic actions by storing target system reactions under defined boundary conditions before actual operation. These pre-stored reference reactions enable proactive fault detection during autonomous operation without requiring driver intervention or complex real-time diagnostic infrastructure.
Solution Approach 2:
The system creates copies of normal system reactions under defined boundary conditions and stores them as target reactions. By comparing actual system reactions against these copied reference patterns, the system can detect deviations indicating faults without requiring additional sensors or complex diagnostic hardware.
2Loss of information
If the vehicle drives while maintaining boundary conditions to collect diagnostic data, then diagnostic information can be gathered, but the standard mode of operation is disturbed and driver freedom is restricted
Solution Approach 1:
Instead of collecting diagnostic data during normal operation by imposing boundary conditions on the driver, the system performs preliminary diagnostic actions by storing target reactions under defined boundary conditions during system setup or maintenance phases. This eliminates the need to restrict driver freedom during actual operation while still enabling comprehensive diagnostic data collection.
3Device complexity
If rigid limit values are used for fault detection, then the diagnostic system is simple to implement, but manufacturing and assembly variances cause false fault indications
Solution Approach 1:
The system changes from using rigid, fixed limit values to using flexible target system reactions that are stored under defined boundary conditions. These target reactions inherently account for manufacturing and assembly variances by capturing the actual system behavior within normal tolerances, thereby improving fault detection accuracy without significantly increasing system complexity.
4Device complexity
If no direct sensor is provided downstream for fault detection, then the system structure remains simple, but mechanical or pneumatic faults cannot be detected
Solution Approach 1:
Instead of installing additional direct sensors for fault detection, the system creates copies of normal system reactions from existing sensor data and stores them as target reactions. By comparing actual reactions against these copied reference patterns, the system can detect mechanical or pneumatic faults using the existing sensor arrangement without adding complex diagnostic hardware.
Data Source
AI summary
A diagnostic method performs a self-diagnosis of an electronically controllable pneumatic braking system for a commercial vehicle. The method includes: receiving a learning signal about the braking system; in response to the reception, putting the braking system into a learning mode and performing the steps: performing a predetermined first activity of the braking system with the commercial vehicle stationary or moving; detecting via a sensor arrangement a first learning system reaction of the braking system in response to the execution of the first activity; and storing the detected first learning system reaction as the first target system reaction in a memory. A braking system and a computer program are also disclosed.